English

Testing network correlation efficiently via counting trees

Statistics Theory 2022-04-05 v2 Machine Learning Statistics Theory

Abstract

We propose a new procedure for testing whether two networks are edge-correlated through some latent vertex correspondence. The test statistic is based on counting the co-occurrences of signed trees for a family of non-isomorphic trees. When the two networks are Erd\H{o}s-R\'enyi random graphs G(n,q)\mathcal{G}(n,q) that are either independent or correlated with correlation coefficient ρ\rho, our test runs in n2+o(1)n^{2+o(1)} time and succeeds with high probability as nn\to\infty, provided that nmin{q,1q}no(1)n\min\{q,1-q\} \ge n^{-o(1)} and ρ2>α0.338\rho^2>\alpha \approx 0.338, where α\alpha is Otter's constant so that the number of unlabeled trees with KK edges grows as (1/α)K(1/\alpha)^K. This significantly improves the prior work in terms of statistical accuracy, running time, and graph sparsity.

Keywords

Cite

@article{arxiv.2110.11816,
  title  = {Testing network correlation efficiently via counting trees},
  author = {Cheng Mao and Yihong Wu and Jiaming Xu and Sophie H. Yu},
  journal= {arXiv preprint arXiv:2110.11816},
  year   = {2022}
}